Canadian Homeless Shelter System

    Eradicating Homelessness has long been a “dream” of activists and non-activists alike. Its a problem that resonates with most Canadians today since many are facing similar precarious housing situations. The lack of safe and affordable housing is a critical social and economic issue that most political leaders have largely ignored. Perhaps the biggest problem in addressing homelessness is the lack of transparency - we simply do not know how many people are homeless in Canada. Although homelessness is arguably one of the most important social issues of our times, current homeless “counting” strategies appear to be largely ineffective. Access to accurate and timely information about the number of homeless people is the only rational and meaningful way of accurately accessing the success or failure of any proposed solution. Single day census data carried out in a select number of communities is not a very effective or ideal way of monitoring homelessness across the country.
    It’s therefore not surprising that some of the latest government statistics on homelessness, such as National Service Provider List (2022)^1,2 and Infrastructure Canada’s Homeless Shelter Capacity Point-in-Time Counts (2022)^1,3, paints a very bleak picture of the problem. Besides the increasing rates of homelessness across the country, the “snap-shot” approach to monitoring the homeless population is, as alluded to above, neither desirable nor prudent. According to the NSPL data (Fig. 1) there were approximately 30,402 permanent beds provided by 1,120 emergency and transitional housing shelter services in different parts of Canada in 2022. By comparison Infrastructure Canada’s “Point-in-Time” (PiT) data (2022) shows that in some 72 communities and regions across Canada there were an estimated 39,713 beds provided by 1,725 shelters (Fig. 2). The discrepancies between these two 2022 lists suggests that we also don’t have an accurate measure of the number of shelter service providers in the country.

Figure 1: 2022 National Service Provider List (NSPL). The graph shows the total number of permanent beds provided by various emergency and transitional homeless shelters from each Canadian province and territory. According to the Homelessness Policy Directorate of Infrastructure Canada there were 30,402 beds provided by 1,120 shelters in 2022.^1,2

Figure 2: Provincial and Territorial Homeless Shelter Capacity PiT Count Data. The graph shows total number of shelters and shelter beds for each province and territory over a seven year period (2016 - 2022). These counts were carried out on a single day (i.e. “Point-in-Time” or PiT) and thus only provide a brief glimpse of the homeless situations within each of the communities that took part in this census. Overall there were 72 communities and regions that participated in the last three annual PiT counts, and only 67 municipalities that took part in all seven PiT counts. These enumerations were based on three so called “core” locations: (i) Unsheltered locations (i.e. streets, alleys, parks, transit stations, abandoned buildings, encampments, vehicles, ravines, and other outdoor locations), (ii) Sheltered locations (i.e. emergency shelters, extreme weather shelters, Domestic Violence shelters, hotel/motel shelter spaces), and (iii) Transitional housing (i.e. longer-term housing spaces that help individuals transition to more secure housing). Counts based on the latter two locations were supplied by shelter service administrators, while counts carried out in unsheltered locations were generated by survey. Counts generated by survey also included: (i) individuals who were observed to be experiencing homelessness outside, but were unable to complete the survey; (ii) people within health and correctional facilities; and (iii) people who staid with other people due to the lack of a permanent residence (e.g. “couch surfacing”).^1,3

Figure 3: National Homeless Shelter Capacity PiT Data (2016 - 2022). The graph shows the overall national “Point-in-Time” (PiT) census of (i) Domestic violence, (ii) Emergency, and (iii) Transitional homeless shelters across Canada for the indicated time period. The latest counts (2022) show that there were a total of 39,713 shelter beds provided by 1,725 shelter service providers in 72 communities and regions from across Canada. Three different types of shelters make up these totals, including: (i) 518 Emergency Shelters that account for 18,467 beds; (ii) 602 Transitional Housing spaces that account for 11,935 beds; and (iii) 605 Domestic Violence Shelters that account for 9,311 beds.^1,3

GTA Shelter System - Population in Crisis

Homelessness, as defined by Canadian Observatory on Homelessness (COH, formerly the Canadian Homelessness Research Network):

“describes the situation of an individual, family or community without stable, permanent, appropriate housing, or the immediate prospect, means and ability of acquiring it. It is the result of systemic or societal barriers, a lack of affordable and appropriate housing, the individual/household’s financial, mental, cognitive, behavioural or physical challenges, and/or racism and discrimination.”⁴

The negative social and economic consequences of homelessness in Canada are enormous. Homelessness has in fact flourished in past few decades, largely because of a lack of government leadership. The current shelter strategy that “warehouses” the poor, keeping them “out-of-sight” and “out-of-mind”, has failed miserably. These types of “band-aide” solutions simply do not address the root causes of homelessness.

“In some ways, it is simple: Canada can choose to invest an average of $2 billion annually to save $4 billion, and bolster urban economies, keep hundreds of thousands of Canadians healthy and productive as well as ensure productive futures. Or we can simply continue with the status quo and hope that things correct themselves, although after nearly 15 years since the original 1993 federal retreat from housing, Canada has succeeded mainly in increasing homelessness.”⁵

In fact, the current status quo is “actually really expensive”.⁶ It is far cheaper and more cost effective to provide homeless people with real housing, as apposed to “band-aide” solutions like shelters, hostels, warming centres and other like-minded temporary emergency services. Current evidence based research clearly “points to the fact that if we do things differently, we not only achieve better social outcomes, but we also save money.”⁶ Homelessness is estimated to cost the Canadian economy 7.05 billion dollars on an annual basis.

The chronic lack of safe and affordable housing, along with the growing economic divide between rich and poor, needs to be addressed by all levels of governments. The current scatter-shot approach to mitigating or eliminating homelessness and poverty, which are inextricably linked, has produced little in the way of real positive social/economic change.

“Not repeating past mistakes should at least be a first line of defence…[such as]curtailed social assistance…punitively low minimum wage rates…tax incentives that ignore renters and those on social assistance…cutbacks to social services and health.”⁵

About Toronto’s Shelter System:
Information in this report is based on a number of Shelter and Homelessness records that are freely available for download from the City of Toronto’s Open Data Portal. The following is a brief overview of the main datasets analyzed by our group, the Voices 4 Unhoused Liberation.
(i) Central Intake (CI)⁷ is the entry point for persons in need of emergency shelter services. People able to access CI’s 24/7 telephone line (416-338-4766 or 1-877-338-3398) can potentially receive a referral to an emergency shelter or other overnight accommodations. This dataset contains records of calls received by CI, and the outcomes of those calls (e.g. # unique individuals who contact CI, but were not matched to a shelter space each day). All of this information is collected by the city of Toronto’s Shelter Management Information System (SMIS). Callers not matched to an available space that suits their needs remain in the queue until they can be provided a referral, or until the closeout process at the end of the night at 4:00 AM. These daily call counts and outcomes (i.e. SMIS assigns codes for various outcomes) are collected and reported on a monthly basis (i.e. monthly updates are published on the 15^th day of the following month). CI uses 13 distinct wrap-up codes to classify the daily calls they receive. Unfortunately, these codes only provide information about the volume and types of calls, and do not contain information about the outcomes of said referrals (Table 1).
(ii) Toronto Shelter System Flow (TSSF)⁸ data is compiled on a monthly basis by the city of Toronto’s Shelter Management Information System (SMIS). These records show the number of people entering (intake) or leaving (outtake) the shelter system. Some of the terms used in these records are listed in Table 2. Unfortunately, the current SMIS data is limited since it does not include any information about homeless people sleeping outdoors, or homeless people who do not use city funded services that are apart of SMIS. Although it is difficult to assess the exact number of homelessness people within the city, the most recent Street Needs Assessment (2021) suggests that ~18% of city’s homeless population are not included in the current data. This is likely an under estimate of its “true” size given the severity of the current housing and homelessness problems within the region.⁸
(iii) Daily Shelter Overnight Service Occupancy (DSOSO)⁹ records maintained by SMIS is compiled on a daily basis. It records the number of people who use overnight shelters and allied services within the GTA. The data includes the identify of program operators, as well as the locations and nature of the services they provide (Table 3).⁶
    Unfortunately there is little information about the amount of funding the city receives and allocates for these different homeless programs and services. The DSOSO data only mentions Funding Capacity within the context of the “number of beds or rooms that a program is intended [funded] to provide”.⁹ Although there may be a number of reasons why beds or rooms may be temporarily out of service (e.g. repairs, renovations, outbreaks, or pest control) they are not reported. These records also mention Actual Capacity, which refers to the “number of beds or rooms in service and showing as available for occupancy in the SMIS at time of reporting.” This latter measure is touted as being “a more effective capacity measure to assess program occupancy rates”. However, the number of unavailable rooms or beds (Note: # UNAVAILABLE BEDS = Capacity Funding Bed - Capacity Actual Bed) also tells us the number of rooms/beds the current system is funding but not using on a daily basis. Unfortunately there appears to be little oversight of the service providers with regards to reducing or eliminating bed/room “unavailability”, and the draconian nature of the service restrictions they impose on clients.

Toronto CIC and TSSF Data

Figure 4. Toronto Central Intake Daily Call Totals. The plotted data, which was taken from the “Central Intake Call Wrap-Up Codes” dataset, provides daily counts of calls answered by Central Intake Centre (CIC). The calls are classified by a caseworker using 13 distinct “wrap-up” codes (i.e. describes general nature of the call). These daily SMIS records are manually updated on a monthly basis. Although the “wrap-up” codes provide information related to the volume and type of calls received, it does not track requests made by unique individuals, nor the ultimate outcomes of referrals.⁷ Looking at the plotted data it is evident that there has been an overall increase in the number of calls to CIC. Unfortunately it appears that few callers were able to successfully obtain a referral for an overnight sleeping or resting space (colour coded). This low success rate has been particularly pronounced within the last 10 months.

Figure 5. Toronto Central Intake Total Numbers of Daily Unmatched Callers. The plotted data, taken from the monthly updated “Central Intake Service Queue” dataset, provides daily counts of unique individuals who contacted the CIC to access an emergency shelter, but were not matched to a shelter space. The SMIS service queue caller data details when a bed was requested for a caller seeking a shelter space. An unmatched caller (i.e. person not given a referral to an available space that suits their needs at the time of their call) remains in the queue until either they were are provided a referral, or until “close-out” (i.e. 4:00 AM, end of the night). The data only contains three features, namely: (i) date of the call, (ii) total number of Unmatched Callers, and (iii) whether the caller phoned once (i.e. Single Caller) or multiple times (i.e. Repeat Caller).⁷ Looking at the plotted data it is evident that there has been a steady increase in the number of callers not receiving a referral to an overnight sleeping space or safe shelter despite repeated calls.

Figure 6. Toronto Shelter System Flow Data Total Numbers. This plot shows the total number of “actively” homeless people according to SMIS records. The TSSF data is only updated on a monthly basis (e.g. recent March 2024 data was updated on April 15^th). The data is only based on those people who have used overnight services that are monitored by SMIS (i.e. excludes people who sleep outdoors or who use non-SMIS homeless services). The changing width of the plotted line reflects the changing dynamics of the homeless population over this 6 year period. Clearly the number of homeless people remains high within the GTA. The recent all time high of 11,173 “actively” homeless people for Feb/2024 is currently ebbing somewhat (10,833 Mar/2024). This is perhaps due to the warmer weather as many homeless people often escape the shelter system for safer environs. Although far from complete the TSSF records appear (for some unknown reason) to be more extensive than the DSOSO records since the latter dataset reports fewer homeless people (i.e. 9,748 for Mar 1^st, 9,653 for Mar 15^th, and 9,687 for Mar 31^st 2024). This discrepancy in the number of homeless people is a bit disconcerting given that both datasets are reportedly based on the same SMIS records.

Figure 7. Demographics of TSSF dataset. The above plot shows the number of “actively” homeless people classified as either (i) Chronically homeless, (ii) a member of a homeless Family, (iii) Indigenous, (iv) a non-refugee, (v) a refugee, (vi) a single adult (> 24 yrs), or a (vii) Youth (16-24 yrs) who is not a member of an homeless family. Like Figure 6, the changing width of the plotted line reflects the changing dynamics of each homeless group. Currently (Mar 2024) the largest homeless demographic group is single adults (Note: 7266 single adults, or 67.1%, compared to 2555 family members, or 23.6%, and 1012 youths, or 9.3%). Adults (> 24 yrs) also make up ~77% (4835/6288) of the chronic homeless population. The homeless refugee population (5805, or 53.6%) slightly out numbers the homeless non-refugee population (5028, or 46.4%).

Figure 8. Demographics of TSSF - In/Out Shelter Flow. The above plot shows the number of people for each of the indicated demographic groups who have either entered or exited the shelter system since 2018. The plotted number of shelter occupants is an aggregate of those who are new to the shelter system and those who have previously left the shelter system and subsequently returned to the shelter system. The above plots also show the number of individuals who have left the shelter system for permanent housing (i.e. represented by the colour of the plotted line, see colour coded number scale). The above plot, like the one shown in Figure 7, contains the following demographic groups: (i) Chronically homeless, (ii) Family members, (iii) Young people (16-24 yrs), (iv) Indigenous people, (v) single adults (> 24 yrs), (vi) refugees, and (vii) non-refugees. Overall, these plots clearly show that the number of homeless people moving out of the shelter system and into permanent housing are relatively few in number. Indigenous and Young (< 24 yrs) people appear to be the two groups most affected by the lack of access to permanent housing over the past 6 years. Moreover, there is another relatively large group of former shelter occupants, termed “Inactive” or “Moved to Inactive”, that are not represented in these plots. According to TSSF records these are people who were using the shelter system up until three months ago, but have apparently fallen off the “radar” (i.e. housing or homeless status currently unknown).

Toronto DSOSO Data

Figure 9. Total Shelter Population within the Greater Toronto Area (GTA). The above plot shows the total number of daily shelter users for the period spanning 2021-01-01 to 2024-04-21. The number of shelter occupants have increased by more than 63% during this time.

Figure 10. Shelter Population per GTA region. The plot shows the total number of shelter users within different parts of the GTA (i.e. Toronto, North York, Etobicoke, Scarborough, Vaughan) for the past ~40 months (2021-01-01 to 2024-04-21). The data reveals that the vast majority of shelter users are currently being housed within Toronto and North York (see Bar graph in Figure 10).

Figure 11. Current Shelter Populations per GTA Region. This bar graph shows the current (2024-04-21) number of shelter occupants within each of the designated GTA regions. Toronto (68.5%) and North York (18.3%) shelter the majority of homeless people, while Scarborough (8.7%), Etobicoke (3.9%), and Vaughan (0.6%) collectively house less than 14% of the homeless population.

Figure 12. Homeless Demographics within the GTA Region. The above graph looks at some homeless demographics within the GTA. The number of individuals within each of the designated homeless groups (i.e. men, women, families, Youth and Mixed adults) have rapidly increased over the last 3 years. Also, the size of the Mixed adult population has fluctuated quite dramatically over the past two years. Possible reasons for this may be due to the fluctuating number of designated “mixed adult” shelter spaces. For example, some service providers, like Homes First, are unilaterally transitioning some of their shelter spaces (e.g. Delta Hotel Shelter, Scarborough) to a select homeless population (i.e. refugee only) at the expense of other homeless people (i.e. non-refugees mixed adult population). Note that the fate of the Delta non-refugee homeless population has largely being determined by Ward 22 councillor Nick Mantas, who’s NIMBY motion to rid the Delta shelter of non-refugee residents was passed by Toronto City Council meeting back in Feb 2024.

Figure 13. Homeless Population Program Model. The above graph looks at the designated Emergency and Transitional shelter programs within the GTA. According to City of Toronto’s Housing + Homelessness Service Glossary (2019)¹⁰ the Emergency Shelter Program constitutes a homeless shelter program that individuals and families can access without a referral. By contrast a Transitional Shelter Program constitutes a homeless program that individuals and families can access only with a referral. The plotted DSOSO data clearly shows that only ~1 out of every 10 homeless individuals have been able to access a referral. This poor referral rate is also evident from the data plotted in Figure 4 and Figure 5.

Figure 14. Homeless Overnight Service Types within the GTA. Homeless individuals within the GTA can access different types of overnight services, including: (i) Shelters, (ii) 24-Hour Respites, (iii) Motels/Hotels, (iv) Interim Housing, (v) Warming Centres, (vi) 24-Hour Women’s Drop-ins, and (vii) Isolation/Recovery Sites. There are a few other Service Types offered, but they are ill-defined and accessed by very few people. As of April 21^st 2024 the vast majority of Service Types are emergency type Shelter (4845, or 48.9%) and Motel/Hotel (4461, or 45%) services. Unfortunately, services that help homeless people transition to safer and more affordable permanent housing are largely non-existent within the Toronto Shelter system.

Figure 15. Homeless Program Area within the GTA. Homeless programs can be categorized as either: (i) Base Shelter, (ii) Base Program - Refugee, (iii) Temporary Refugee Response, (iv) COVID-19 Response, or (v) Winter Response. Generally they can be classified as either regular (i.e. year-round) or temporary programs. The rising numbers of homeless refugees and along with the long-term health and economic impacts of the COVID-19 pandemic appears to be driving the growing homeless population within the GTA. Even the cyclic Winter Response program, which normally spikes during the cold season, experienced a considerable rise in the number of homeless people during the last winter season.

Figure 16 & 17. Homes First’s Delta Hotel Shelter. These two graphs show the number of rooms at the Delta Hotel Shelter that are either occupied (Figure 16) or unoccupied (Figure 17). Although the number of occupants have been steadily decreasing between Feb 5^th to April 4^th, the recent influx of refugees has rapidly corrected these falling numbers. The rapid changes in the number of people sheltered at the Delta Hotel over the past 2 months is also reflected in the number of unavailable rooms at the Delta shelter (Figure 17). The fitted Loess (LOcal regrESSion) curve is fairly well defined (i.e. residuals are fairly constant over the chosen range of x values), which suggests “real” trends in the data (e.g. rapid increase in # unavailable rooms between Feb 5th and April 4th, followed by a quick reversal of this trend between April 4th to present). Nevertheless, it is important to note that the overnight Delta Hotel Shelter (Program ID 16011) supervised by Homes First is, according to DSOSO records, a designated mixed adult (Sector) shelter that falls under the emergency COVID-19 Program. So it is by definition not a refugee-only shelter space. According to Homes First, which runs many other additional shelter and supportive housing sites, their mission is to “develop and provide affordable, stable housing and support services to break the cycle of homelessness for people with the fewest housing options”. However, despite the continuing rise in the number of homeless persons in Toronto (Fig. 6,9,12,13), as well as the recent influx of Federal dollars ($162 million) to support the Toronto shelter system, Homes First has unilaterally decided to transition the Delta Hotel Shelter to a refugee only space (i.e. no consultations with residents). Apparently, in accordance with the wishes of local Ward 22 councillor Nick Mantas, the shelter is actively trying to move or evict people from the safety of this shelter. Many have already faced “service restrictions” without any appeals process. Even the newly implemented appeals process typically results in the eviction of the client at the end of the restriction period. It is important to note that during the time it takes to resolve the appeal process the client is removed from the shelter and forced to find accommodations on their own (i.e. left homeless). Homes First’s refusal to consult with Delta residents about the fate of the shelter, one that directly impacts their immediate and future well being, is both imprudent and overtly punitive in nature. Many residents are afraid that they will be forced to move back onto the streets, or into some unsafe open shelter space. Many have already been severely traumatized (i.e. physically and verbally assaulted, victims of theft, witness friends and people die) by many of the current shelter providers. The lack of transparency about the fate of the Delta shelter has only heightened the resident’s fears, anxiety and distrust of Homes First. This situation has only been exacerbated by the lack of available housing workers at the Delta shelter. The current drive to transition the Delta Hotel to a refugee-only shelter, at the expense of non-refugees, is perplexing given the recent influx of Federal dollars into the Toronto shelter system. Shelter spaces should be inclusive by nature, and not resort to punitive measures that divide and further traumatize the most marginalized members of our society.

References

1. Havi Echenberg and Laura Munn-Rivard (2020). Defining and Enumerating Homelessness in Canada. Publication No. 2020-41-E, Parliamentary Information and Research Service, © Library of Parliament, Ottawa, Canada 2020.

2. The National Service Provider List - The National Service Provider List - 2022.

3. Statistics Canada. Table 14-10-0353-01 Homeless Shelter Capacity in Canada from 2016 to 2022, Infrastructure Canada. DOI: https://doi.org/10.25318/1410035301-eng

4. Gaetz, S.; Barr, C.; Friesen, A.; Harris, B.; Hill, C.; Kovacs-Burns, K.; Pauly, B.; Pearce, B.; Turner, A.; Marsolais, A. (2012). Canadian Definition of Homelessness. Toronto: Canadian Observatory on Homelessness Press. www.homelesshub.ca/homelessdefinition

5. Laird, Gordon (2007). SHELTER - Homelessness in a growth economy: Canada’s 21st century paradox. Report for the Sheldon Chumir Foundation for Ethics in Leadership.

6. Gaetz, Stephen (2012). The Real Cost of Homelessness: Can We Save Money by Doing the Right Thing? Toronto: Canadian Homelessness Research Network Press.

7. City of Toronto (2024). Central Intake Calls (CIC).

8. City of Toronto (2024). Toronto Shelter System Flow (TSSF).

9. City of Toronto (2024). Daily Shelter and Overnight Service Occupancy and Capacity (DSOSO).

10. City of Toronto - Housing Stability Service System Overview (2019). Housing + Homelessness Service Glossary.